EVOLUTION & DEVELOPMENT

3:1, 18–23 (2001)

©

BLACKWELL SCIENCE, INC.

18

Evolution and development of gastropod larval shell morphology:

experimental evidence for mechanical defense and repair

Carole S. Hickman

Department of Integrative Biology, University of California, Berkeley, CA 94720-3140, USA

Correspondence (email: caroleh@ucmp1.berkeley.edu)

SUMMARY

The structural diversity of gastropod veliger lar-vae offers an instructive counterpoint to the view of larvalforms as conservative archetypes. Larval structure, function,and development are fine-tuned for survival in the plankton.Accordingly, the study of larval adaptation provides an impor-tant perspective for evolutionary-developmental biology asan integrated science. Patterns of breakage and repair in thefield, as well as patterns of breakage in arranged encounterswith zooplankton under laboratory conditions, are two power-

ful sources of data on the adaptive significance of morphologicaland microsculptural features of the gastropod larval shell. Shellsof the planktonic veliger larvae of the caenogastropod

Nas-

sarius paupertus

[G

OULD

] preserve multiple repaired breaks,attributed to unsuccessful zooplankton predators. In cul-ture, larvae isolated from concentrated zooplankton samples

rapidly repaired broken apertural margins and restored the

“ideal” apertural form, in which an elaborate projection or“beak” covers the head of the swimming veliger. When indi-viduals with repaired apertures were reintroduced to a con-centrated mixture of potential zooplankton predators, therepaired margins were rapidly chipped and broken back. Theprojecting beak of the larval shell is the first line of mechanicaldefense, covering the larval head and mouth and potentiallythe most vulnerable part of the shell to breakage. Patterns ofmechanical failure show that spiral ridges do reinforce thebeak and retard breakage. The capacity for rapid shell repairand regeneration, and the evolution of features that resist orretard mechanical damage, may play a more prominent rolethan previously thought in enhancing the ability of larvae tosurvive in the plankton.

INTRODUCTION

The union of evolutionary and developmental biology hasgenerated new interest in the origins, evolution, and rolesof larvae in metazoan life history (Hall and Wake 1999;Wolpert 1999; Nielsen 2000). It has been easy for biology asa whole to view development primarily as a program for build-ing the adult body plan. However, the larva may be the de-scendant of the original metazoan, and metamorphosis andthe adult phenotype an evolutionary add-on (Nielsen 1995,2000). Whether larvae are primary or secondary in metazoanlife cycles, they must be targets of natural selection to func-tion well in an environment that differs markedly from thoseof adults. This is one of a series of articles investigating thestructural diversity of gastropod larval shells in both an evo-lutionary and a developmental context (Hickman 1995,1999a,b).

The objective of this research is to direct attention to thelarval phenotype as a target of selection for larval function.It provides experimental evidence for both structural anddevelopmental attributes that promote larval survival in theplankton. Structural adaptations include the larval “beak,”

a prominent projection of the growing shell margin directlyover the head and mouth of the swimming larva, and fea-tures of the beak that retard mechanical breakage and chip-ping. Developmental adaptations include not only a mech-anism for repair of broken larval beaks, but also a larvalsecretory mechanism that further reinforces the repairedshell.

Swimming and feeding veliger larvae of marine gastro-pods differ in many aspects of shell morphology and bodyplan from that of post-metamorphic benthic juveniles andadults. Exclusively larval structures such as the velumclearly have evolved under selective pressure for perfor-mance in swimming and feeding. However, the adaptivesignificance of larval shell form and structure is still largelyunexplored. If larval mortality results primarily from pred-ators swallowing larvae whole, as frequently assumed(Jägersten 1972; Vermeij 1993), then natural selectionwould have little opportunity to produce defensive struc-tures.

The assumption that the entire gastropod larva is alwaysat risk underlies the only experimental test of an innova-tion called larval torsion, which is alleged to have an anti-

Hickman

Gastropod larval shell morphology

19

predatory adaptive significance. Larval torsion, which twiststhe head and foot by 180

8

relative to the shell, mantle, andvisceral mass, is a developmental event with structural andfunctional consequences. Garstang (1928, 1929) was thefirst to hypothesize that torsion would have an immediateadaptive advantage for the gastropod larva by allowing it towithdraw its head and velum rapidly into the shell when un-der attack by a predator. Pennington and Chia (1985) wereunable to find evidence that torsion conferred any protectionto veliger larvae of

Haliotis kamtschatkana

[

Jonas

]. How-ever, the predators in their study were ones that ingestedwhole larvae, and potential for torted larvae to have in-creased rates of survival under sublethal attack was nottested.

Does the larval shell itself have a protective function?Does it have mechanical and developmental attributes thatpromote survival to settlement and metamorphosis? The or-igin of biomineralized exoskeletons, near the beginning ofthe Cambrian Period, is most commonly interpreted as a pro-tective innovation (Stanley 1976; Vermeij 1987, 1990;Bengston 1994). However, the evidence for protective func-tion is based exclusively on the frequency of shell repair onadult shells in the fossil record (Vermeij et al. 1980, 1981;Schindel et al. 1982; Ebbestad and Peel 1997).

There are three recent lines of evidence leading to theconclusion that recurring structural features of gastropod lar-val shells do contribute to mechanical defense (Hickman1999a, b). The first is observation of repairs of sublethalbreakage on shells of larvae collected from the Hawaiianplankton (Hickman 1999a, b). The second line of inferenceis from application of the Paradigm Method, developed bypaleontologists for the analysis of structures of extinct fossilorganisms (Rudwick 1964), to larval shells in 21 families ofcaenogastropods from the Hawaiian plankton (Hickman1999b). A set of four recurring features fit ideal engineeringsolutions for retarding breakage in thin ceramic materials.The third line of inference, supporting the conclusionsreached using the Paradigm Method, is the observation thatthe same set of four features deters mechanical damage in ar-tificial “attacks” on larval shells in the laboratory (Hickman1999b).

This article reports the results of a fourth line of evidence:results of laboratory tests in which larvae in culture respondto real encounters with zooplankton. The approach is quali-tative and empirical, featuring Scanning Electron Microscope(SEM) observations of larval shells of a species from the Ha-waiian plankton in which chipping and breakage are fre-quent, not only on the growing margins of shell but also aspreserved multiple injuries that had been repaired routinelyby larvae during growth. Mechanical damage occurred onlyin the presence of potential zooplankton predators and not inisolation, implicating a possible source of the selection pres-sure on larval shell morphology and its secretory mechanism.

MATERIALS AND METHODS

Larval snails and sampling

All of the larvae of the caenogastropod

Nassarius

paupertus

[

Gould

]used in this study were recovered from a single plankton tow im-mediately offshore from the reef at Kewalo Basin, southeasternOahu Island, Hawaii (12

8

17

9

30

0

N, 157

8

51

9

30

0

W) on 13 October1993. Veliger larvae were removed and sorted over an 8-h periodby repeated swirling of the sample in a bucket containing approxi-mately 10 l of seawater. Veligers were caught up disproportionatelyin the vortex and settled disproportionately early at the base of thevortex, where they were removed with a turkey baster to a fingerbowl of 0.45

m

m Millipore-filtered seawater for fine-sorting. Theprocedure was repeated until veligers were no longer present. Toavoid mechanical damage to larval shells from handling, all trans-fers were accomplished using a pipette with an opening diameterexceeding that of the larva.

Larvae of

Nassarius paupertus

were placed with a pipette ini-tially into a finger bowl containing 500 ml of 0.45

m

m Millipore-filtered seawater. To retard evaporation while permitting oxygenexchange, finger bowls were covered with plastic film for the dura-tion of the study.

Baseline morphological assessment

To assess the condition of shells as they came from the plankton, 10actively swimming veligers were removed on the day of collectionand immediately placed with pipette into distilled water where theyremained for 5 min while animals retracted fully into their shells.The distilled water treatment also eliminated the possibility of for-mation of salt crystals on the shell during drying.

Shells were transferred subsequently to 70% ethanol where theywere manipulated for several minutes with soft sable brushes to re-move all visible mucus and particulate material from the shell sur-face. Ultrasonic cleaning was not used because of the possibility ofmechanical damage to the shells. From the alcohol, shells weretransferred with a sable brush to lint-free weighing paper and airdried. Air-dried specimens were mounted on aluminum stubs usingsilver paste. Specimens were then coated with gold/palladium alloyand examined with the Hitachi S-800 Field Emission ScanningElectron microscope at the Biological EM Facility of the Universityof Hawaii’s Pacific Biomedical Research Center. All specimenswere viewed at an accelerating voltage of 10 kV and at a workingdistance of 15 mm. All micrographs were recorded on PolaroidType 55 Positive/Negative film (Polaroid, Cambridge, MA, USA).

Morphological assessment of cultured larvae

Ten larvae were cultured in each of four finger bowls, each contain-ing 300 ml of filtered seawater. Larvae in each treatment were of-fered the same food mixture of the diatom

Chaetoceros gracilis

, thechrysophyte

Nannochloropsis occulata

, and the Tahitian strain ofthe flagellate

Isochrysis galbana.

Seawater was changed every 48 h,and bowls were covered with plastic wrap to retard evaporation. Tocharacterize the growth and repair of shells in culture, and in isola-tion from potential predators, larvae were examined daily and notesand drawings were made. On the sixth day, 10 individuals were re-moved and prepared for SEM using the same procedures as in thebaseline assessment.

20 EVOLUTION & DEVELOPMENT

Vol. 3, No. 1, January–February 2001

Predation trials

To determine whether repaired and fresh apertural margins are me-chanically damaged by zooplankton predators, another planktontow was obtained on 20 October 1993, and 10 healthy swimmingveliger larvae from the cultures were placed in each of two 500 mlbeakers of seawater along with samples of mixed concentratedzooplankton, a diverse assemblage in which crustacean larvae werepredominant. A bubbler was placed in each beaker for oxygenation,and at the end of 12 h, the

Nassarius

larvae were removed and pre-pared for SEM using the same procedures as in the baseline and cul-ture assessments. As a control, 10 healthy swimming veligers alsowere placed in a third 500 ml beaker with a bubbler, but without theaddition of zooplankton.

RESULTS

Baseline morphological assessment

Several observations of veliger larvae during initial sortingfrom plankton tows are pertinent to this study. The first isthat the head, foot, and velum usually were fully retractedinto the shell in the finger bowls at the time of sorting. Thesecond is that larval shells frequently were observed underrepeated “attack” by a variety of organisms, particularly lar-val crustaceans, in the concentrated sample. As soon as lar-vae of

Nassarius

paupertus

were pipetted into finger bowlsof filtered seawater and isolated from potential attack, theyexpanded their velar lobes and begin to swim actively.

Apertural margins of individuals examined immediatelyfollowing isolation from the plankton sample typically ap-peared chipped and broken to varying degrees. In the mostextreme instances, the larval beak was broken back to thelevel of the velar notches. Breakage of the apertural marginof three individuals is illustrated in Figures 1–3. It is clearthat similar breakage has occurred under normal conditions,because growth lines on earlier portions of the larval shells

show previous instances of marginal breakage and its subse-quent repair (see arrow in Fig. 3).

Similar apertural chipping was observed on many otherprosobranch species recovered from plankton samples alongwith

Nassarius paupertus.

Breakage patterns vary with spe-cies, and are closely correlated with inferred degree of calci-fication of the shell, the location of the spiral ornament andkeels that appear to retard breakage, and the degree of termi-nal thickening of the velar notches and beaks (Hickman1999a, b). Larvae of

N. paupertus

are not among those withthe most robust mechanical protection, although they show acomparatively high incidence of repair of breakages duringearlier larval shell growth.

Morphological assessment of cultured larvae

Within 12 h of isolation in culture, larvae of

Nassarius pau-pertus

had begun new growth along the apertural margin,with initial growth either restricted to or concentrated at themargin of the beak. After 36 h, the beak had expanded dra-matically, and at the time of removal of individuals from cul-ture for SEM at 60 h, all of the individuals had grown newbeaks (Figs. 4–7). In each of the illustrated individuals, anarrow marks the irregular broken margin of the shell as it wasremoved from the plankton sample.

Repair is accomplished by a novel sequence of secretoryevents. New growth is initiated on the inside of the brokenshell margin by deposition of layers parallel to old growth.Each subsequent layer thickens the shell inward from the old

Figs. 1–3. Larval shells of three individuals at the time of recoveryfrom plankton sample, illustrating breakage patterns. Fig. 1. Beakis broken and spiral ridges indicate resistance at breaks (arrows).Bar 5 231 mm. Fig. 2. Breakage visible along entire apertural mar-gin. Bar 5 270 mm. Fig. 3. Breakage visible along entire aperturalmargin. Repaired break in the posterior velar notch (arrow) pro-vides evidence of earlier damage in the plankton. Bar 5 270 mm.

Figs. 4–7. Larval shells of four individuals after isolation in cul-ture and repair of damaged apertural margin. On each specimenthe configuration of the broken margin at the time of recoveryfrom the plankton sample is visible as a distinct irregular break,and the newly-grown beak and velar notches have smooth mar-gins. Bars 5 270 mm (Figs. 4–6), 250 mm (Fig. 7).

Hickman

Gastropod larval shell morphology

21

surface (“surficial aggradation”) at the site of damage. Fol-lowing reinforcement of the broken shell margin, accretion-ary growth (“marginal accretion”) is resumed, entirely off ofthe new shell material. No new material is added to the mar-gin of the break, resulting in a stepped appearance to the re-pair. Ornamentation on the new growth typically differs fromthe ornamentation terminated by the breakage (Figs. 4–7).There is little or no continuity across the break, althoughthere may be recognizable elements of the program on eitherside of the disruption.

The behavior of repaired individuals in culture suggestedan ability to respond protectively to potential predators.While under observation, swimming veligers were less sen-sitive to changes in lighting and to mechanical disturbance ofwater in culture dishes than are many other prosobranch ve-ligers. However, direct physical contact with a swimmingveliger via the tip of a sable brush caused it to retract the ve-lum. Under repeated contact by the bristles, animals were de-terred from re-extending the velum and sank to the bottom ofthe culture dish. With removal of the bristles, the velum wasre-extended and swimming resumed.

Predation trials

When larvae with pristine beaks were transferred from cul-ture into beakers with concentrated fresh zooplankton, thebeaks were broken back rapidly (Figs. 8–10). In the treat-ment that was left for 12 h, two of the 10 individuals disap-peared and the remaining eight were all dead on the bottomof the beaker, with animals completely retracted. After 4 h,breakage was at a level comparable to that observed in indi-viduals at the time of initial sorting and did not completelyobliterate the growth increments laid down in culture. In Fig-ures 8–10, an arrow marks the original broken area where re-pair began in culture. After 12 h, no breakage was observedin the control treatment.

DISCUSSION

This study reinforces the view that larvae are targets of nat-ural selection. The larval phenotype reflects selective pres-sure for mechanical defense and developmental repair of me-chanically inflicted damage in the course of larval life.

Specifically, the results of this study indicate that larvaeof

Nassarius paupertus

survive encounters with zooplanktonin which the growing apertural shell margin is repeatedlychipped or broken and subsequently repaired. The beak,which is the most forward-projecting portion of the larvalshell margin, is reinforced by spiral ridges or thickeningsthat retard breakage and typically stand out as the most resis-tant points along a broken shell margin. Because the beakcovers the head and mouth of the swimming larva, it is ap-propriately located to form a protective function.

Larval shells that record episodes of sublethal breakageand repair are powerful indicators of past physical encoun-ters in the plankton. They are particularly significant as evi-dence that breakage and repair do occur at natural populationdensities in the field. Laboratory trials are important for doc-umenting that damage occurs in the presence of zooplank-ton. However, it is probable that the breakage illustrated hereis in excess of what normally occurs in the field and is attrib-utable to the amount of time that larvae spent in a concentrateof potential zooplankton predators. Larvae experienced ele-vated zooplankton densities during this study: first, in thecod end of the plankton net; second, during the interval ofsorting; and finally, in the reintroduction trials. Accordingly,I do not infer any correlation between the potential to en-counter predators at natural densities in the water columnfrom observations at elevated densities in the laboratory. Thelaboratory portion of this study characterizes the pattern ofmarginal shell growth and repair in isolation from potentialpredators, and documents a subsequent episode of marginalbreakage that is directly linked to reintroduction of larvae tozooplankton.

Although it is tempting to infer that observed breakagewas a direct result of attempted predation, no such claim ismade. Further research is required to determine the specificnature of the encounters that caused the observed damage,although attempted predation is a strong hypothesis. A briefdiscussion of the ecological perspective on predation in theplankton will help distinguish between questions of larvalecology and questions that await investigation in evolution-ary developmental biology.

An important distinction is that wholesale ingestion oflarvae by large predators is of great importance in the analy-sis of life history strategies as they pertain to evolutionarypressures on adult reproduction, but it does not represent anevolutionary pressure on larvae if there is no way to escape.The conventional wisdom that larval mortality results prima-rily from being swallowed whole by predators (Jägersten

Figs. 8–10. Larval shells of three individuals after reintroductionto a concentrate of mixed live zooplankton. On each specimen theline of initial breakage and repair is still visible (arrows), and thenew damage is least extensive at the beak and has been retardedby spiral elements of beak sculpture. Bars 5 270 mm (Figs. 8, 10);250 mm (Fig. 9).

22 EVOLUTION & DEVELOPMENT

Vol. 3, No. 1, January–February 2001

1972; Vermeij 1993) has retarded the search for functionalsignificance in larval morphology. Larvae become muchmore interesting to evolutionary and developmental biologyif they can survive encounters with predators.

Through studying the ecology of marine invertebrate lar-vae, great strides have been made in understanding processesthat affect recruitment of adult populations. The classic stud-ies of Gunnar Thorson (1936, 1946, 1950, 1966) stimulateda major body of ecological research, which assessed alterna-tive sources of “larval wastage” and strived to quantify therisk of having a planktonic stage of development. The ideathat predators were major consumers of larvae dates fromLebour (1922, 1923), and predation currently is consideredto be a significant (Young and Chia 1987), if not the singlemost important (Morgan 1995), source of natural larval mor-tality. Under laboratory conditions, planktonic invertebratelarvae are consumed by a variety of predators (Morgan 1955).

Experimental studies of the effects of predation on plank-tonic embryos and larvae have focused on instantaneousmortality (see Rumrill 1995 for a review), and therefore havebeen designed to measure the disappearance (equals con-sumption) of individuals from experimental vessels (Pen-nington and Chia 1984; Rumrill and Chia 1985; Rumrill et al.1985) or the presence of prey in the gut contents of predators(Johnson and Brink 1998). Based on observations of the gutcontents of polychaete trochophores, larval polychaetes havebeen implicated as significant predators on bivalve veligers(Lebour 1922; Mileikovsky 1959; Wilson 1982; Yokouchi1991). When Johnson and Brink (1998) tested this hypothe-sis in laboratory experiments at what they estimated to benatural prey densities, they found no evidence of polychaetepredation on bivalve larvae. They suggested that literatureimplicating larval polychaetes as bivalve predators could bethe consequence of “unnatural” cod end predation duringplankton tows. However, they state that natural field densi-ties may be strongly underestimated and that frequencies ofencounter could be high as an effect of plankton patchiness(Johnson and Brink 1998).

The predator-prey size difference is an important factor inthe studies previously cited. For example, in the Johnson andBrink (1998) experiments, the oyster larvae offered as preywere all

,

100

m

m in contrast to predator larvae with bodylengths on the order of 230

m

m. Although these are remark-ably large prey to be ingested whole, the veliger larvae of

Nassarius

used in this study were all more than eight timesthe size of the oyster larvae, and clearly too large to be in-gested whole by the zooplankton predators that caused thedamage illustrated here.

The plankton tows for this study were made in a zone thatpreviously and consistently has yielded large numbers ofplanktonic larvae and a diverse zooplankton assemblage, incontrast to a zone immediately seaward that yields typicallyoceanic plankton and few larvae. Concentration of mero-

plankton into fronts may create arenas in which the probabil-ity of encounter is high. It could be the rule rather than theexception that a long-lived feeding molluscan veliger willenter, either passively or actively, into such an arena at sometime during larval life.

It is surprising that so little attention has been paid to thepotential antipredatory function of larval shell features. In asummary of major sources of larval mortality, selectiveagents, and

possible

adaptations, Morgan (1995) listed onlysix general morphological features that might function to de-ter predators: exoskeletons, shells, spines, setae, spicules,and nematocysts. The gastropod larval shell is not a single,simple structure. Gastropod larval shells are variable in size,overall shape, surface sculpture, apertural configuration, mi-crosculpture, microstructure, degree of calcification, and re-lationship of the velum and larval soft parts to the shell. Theresults of this study reinforce other forms of evidence thatlarval shell beaks, thickened velar notches, peripheral angu-lations, and various kinds of spiral and axial sculpture are ef-fective in reducing mechanical damage to larval shells(Hickman 1999a,b).

CONCLUSION

Analysis of larval structure in functional terms contributes toa balanced evolutionary-developmental perspective on ma-rine invertebrate larvae because it emphasizes larval struc-ture and development as targets of natural selection. Con-trasting perspectives have regarded larvae primarily asdispersal mechanisms (Scheltema 1977, 1986), “packages ofgenes dispersed away from each other and away from theparents” (Hart and Strathmann 1995), “devices for turningsmaller eggs into larger juveniles” (Hart and Strathmann1995), “machines for feeding and growing” (Strathmann1987), or “life support system(s) for the imaginal set-asideanlagen within which the adult body plan develops” (David-son et al. 1995). Larvae also have been regarded as represen-tations of ancestors, an idea which, when removed fromHaeckelian logic and cast in a modern developmental per-spective, provides the basis for a viable hypothesis for the or-igins of larvae (Nielsen 1995, 2000). Evolutionary-develop-mental biology provides an arena in which all of theseviewpoints contribute to an integrative understanding of lar-vae and biphasic life history.

Acknowledgments

I am especially grateful to M. G. Hadfield for the opportunity to con-duct this research at the University of Hawaii Kewalo Marine Labo-ratory and for comments on an early version of the manuscript. I amindebted to R. Raff and two anonymous reviewers for comments thathave helped focus the central message. I thank R. Chock and S.Sakata for the plankton tows and T. M. Carvalho and M. Dunlap fortheir assistance in use of the Biological Electron Microscope Lab at

Hickman

Gastropod larval shell morphology

23

the University of Hawaii Pacific Biomedical Research Center. Thisresearch was supported by NSF DEB 9303169 to C. S. H.

REFERENCES

Bengston, S. 1994. The advent of animal skeletons. In S. Bengston (ed.).

Early Life on Earth.

Nobel Symposium no 84. Columbia UniversityPress, New York, pp. 412–425.

Davidson, E. H., Peterson, K., and Cameron, R. A. 1995. Origin of bilate-rian body plans: evolution of developmental regulatory mechanisms.

Science

270: 1319–1325.Ebbestad, J. O. R., and Peel J. S. 1997. Attempted predation and shell re-

pair in Middle and Upper Ordovician gastropods from Sweden.

J. Pale-ont.

71: 1007–1019.Garstang, W. 1928. Origin and evolution of larval forms.

Nature

122: 366.Garstang, W. 1929. The origin and evolution of larval forms.

Rep. Br. As-soc. Adv. Sci.

, Sect. D: 77–98.Hall, B. K., and Wake, M. H. 1999. Introduction: larval development, evo-

lution, and ecology. In B.K. Hall and M.H. Wake (eds.)

The Origin andEvolution of Larval Forms.

Academic Press, San Diego, pp. 1–19.Hart, M. W., and Strathmann, R. R. 1995. Mechanisms and rates of suspen-

sion feeding. In L. McEdward (ed.).

Ecology of Marine InvertebrateLarvae.

CRC Press, Boca Raton, FL, pp. 193–221.Hickman, C. S. 1995. Asynchronous construction of the protoconch/teleo-

conch boundary: evidence for staged metamorphosis in a marine gastro-pod.

Invertebrate Biology

114: 295–306.Hickman, C. S. 1999a. Larvae in invertebrate development and evolution.

In B.K. Hall and M.H. Wake (eds.).

The Origin and Evolution of LarvalForms.

Academic Press, San Diego, pp. 21–59.Hickman, C. S. 1999b. Adaptive function of gastropod larval shell features.

Invertebrate Biology

118: 346–356.Jägersten, G. 1972.

Evolution of the Metazoan Life Cycle.

Academic Press,New York.

Johnson, K. B., and Brink, L. A. 1998. Predation on bivalve veligers bypolychaete larvae.

Bio. Bull.

194: 297–303.Lebour, M. 1922. The food of planktonic organisms.

J. Mar. Biol. AssocU.K.

12: 64–677.Lebour, M. 1923. The food of planktonic larvae. II.

J. Mar. Biol. Assoc U.K.

13: 70–92.Mileikovsky, S. A. 1959. Interrelations between the pelagic larvae of

Neph-tys ciliata

(O.F. Müller),

Macoma baltica

and

Mya arenaria

of the WhiteSea.

Zool. Zh.

38: 1889–1891.Morgan, S. G. 1995. Life and death in the plankton: larval mortality and ad-

aptation. In L. McEdward (ed).

Ecology of Marine Invertebrate Larvae.

CRC Press, Boca Raton, FL, pp. 279–321.Nielsen, C. 1995.

Animal Evolution.

Oxford University Press, Oxford.Nielsen, C. 2000. The origin of metamorphosis.

Evol. & Dev.

2: 127–129.Pennington, T. J., and Chia, F.-S. 1984. Morphological and behavioral de-

fenses of trochophore larvae of

Sabellaria cementarium

(Polychaeta)against four planktonic predators.

Biol. Bull.

167: 168–175.Pennington, T. J., and Chia F.-S. 1985. Gastropod torsion: a test of

Garstang’s hypothesis.

Biol. Bull.

169: 391–396.Rudwick, M. J. S. 1964. The inference of function from structure in fossils.

Brit. J. Phil. Sci.

15: 27–40.Rumrill, S. S. 1995. Natural mortality of marine invertebrate larvae.

Ophelia

32(1–2): 163–198.

Rumrill, S. S., and Chia F.-S. 1985. Differential mortality during the embry-onic and larval lives of northeast Pacific echinoids. In B. F. Keegan andB. D. S. O’Conor (eds.).

Echinodermata.

A. A. Balkema, Rotterdam,pp. 333–338.

Rumrill, S. S., Pennington, T. J., and Chia, F.-S. 1985. Differential suscep-tibility of marine invertebrate larvae: laboratory predation of sand dol-lar,

Dendraster excentricus

(Eschscholtz) embryos and larvae by zoeaeof the red crab,

Cancer productus

Randall.

J. Exp. Mar. Biol. Ecol.

90:193–208.

Scheltema, R. S. 1977. Dispersal of marine invertebrate organisms: paleo-biogeographic and biostratigraphic implications. In E.G. Kauffman andJ.E. Hazel (eds.).

Concepts and Methods in Biostratigraphy.

Dowden,Hutchinson & Ross, Stroudsburg, PA, pp. 73–108.

Scheltema, R. S. 1986. Long-distance dispersal by planktonic larvae ofshoal-water benthic invertebrates among central Pacific islands.

Bull.Mar. Sci.

39: 241–256.Schindel, D. E., Vermeij, G. J., and Zipser, E. 1982. Frequencies of re-

paired shell fractures among the Pennyslvanian gastropods of north-central Texas.

J. Paleont.

56: 729–740.Stanley, S. M. 1976. Fossil data and the Precambrian-Cambrian evolution-

ary transition.

Am. J. Sci.

276: 56–76.Strathmann, R. R. 1987. Larval feeding. In A. C. Giese, J. S. Pearse, and V.

B. Pearse (eds.).

Reproduction of Marine Invertebrates.

Blackwell Sci-entific Publications, Palo Alto, CA and Boxwood Press, Pacific Grove,CA, pp. 465–550.

Thorson, G. 1936. The larval development, growth and metabolism of Arc-tic marine bottom invertebrates compared with those of other seas.

Meddr. Grønland

100(6): 1–155.Thorson, G. 1946. Reproduction and larval development of Danish marine

bottom invertebrates.

Meddr. Komm. Danm. Fisk. - og Havunders. ser.Plankton

4: 1–523.Thorson, G. 1950. Reproductive and larval ecology of marine bottom in-

vertebrates.

Biol. Reviews

25: 1–45.Thorson, G. 1966. Some factors influencing the recruitment and establish-

ment of marine benthic communities.

Neth. J. Sea Res.

3: 267–293.Vermeij, G. J. 1987.

Evolution and Escalation—An Ecological History ofLife.

Princeton University Press. Princeton, NJ, p. 527.Vermeij, G. J. 1990. The origin of skeletons.

Palaios

4: 585–589.Vermeij, G. J. 1993.

A Natural History of Shells.

Princeton UniversityPress, Princeton, NJ, p. 207.

Vermeij, G. J., Zipser, E., and Dudley, E. C. 1980. Predation in time andspace: peeling and drilling in trebrid gastropods.

Paleobiology

6: 352–364.

Vermeij, G. J., Schindel, D. E., and Zipser, E. 1981. Predation through geo-logical time: evidence from gastropod shell repair.

Science

214: 1024–1026.

Wilson, D. P. 1982. The larval development of three species of

Magelona

(Polychaeta) from the localities near Plymouth.

J. Mar. Biol. Assoc.U.K.

62: 385–401.Wolpert, L. 1999. From egg to adult to larva.

Evol. & Dev.

1: 3–4.Yokouchi, K. 1991. Seasonal distribution and food habits of planktonic lar-

vae of benthic polychaetes in Volcano Bay, Southern Hokkaido, Japan.

Ophelia

5: 401–410.Young, C. M., and Chia, F.-S. 1987. Abundance and distribution of pelagic

larvae as influenced by predation, behavior, and hydrographic factors.In A. C. Giese, J. S. Pearse, and V. B. Pearse (eds.).

Reproduction ofMarine Invertebrates.

Blackwell Scientific, Palo Alto and the BoxwoodPress, Pacific Grove, CA, pp. 385–463.